In 1999, Jocelyn M. Hicks, PhD, FRCPath, a past AACC president hospitalized for Guillain-Barre syndrome, was shocked by the amount of blood drawn for her lab tests during her admission. This prompted her to survey 19 community, university, and children’s hospitals about blood-drawing practices. Hicks discovered that most respondents collected 10 times the volume of blood necessary for routine laboratory testing (1). Responses demonstrated 2.5–10 mL of blood was collected for testing that required just 1.5 mL or less.

Little has changed over the ensuing 17 years, despite the fact that modern chemistry analyzers require less than 50 µL of plasma for a routine comprehensive metabolic panel. Recent publications have equated laboratory blood draws as “contemporary bloodletting” and have shown a direct relationship between patients’ clinical fragility and the volume of blood removed for laboratory testing (2,3). Given these findings, it is not surprising that phlebotomy has been implicated as a primary cause of hospital acquired anemia (HAA). One study found that the rate of HAA was approximately 45% in patients older than 18 years admitted to an internal medicine inpatient service (4). A prospective, multicenter observational study of 30 pediatric intensive care units (PICUs) reported a similar rate of 41% (5). HAA also has been associated with increased in-hospital morbidity and mortality.

Due to these concerns, there is a constant push to reduce blood sampling in pediatric patients, and lately, a new impetus to decrease blood draw volumes in adult care settings.  Demand for laboratory testing on smaller blood volumes poses several challenges for laboratories, given the current limitations in automated sample processing and testing equipment. In this article, we address through the lens of our experience in pediatric testing several pre-analytical, analytical, and post-analytical issues that need to be considered when dealing with small sample volumes.

How Low Can you Go?

The amount of blood that can be drawn safely from an individual depends on his or her total blood volume estimated from total body weight (6). Consensus is lacking on how much blood can be obtained safely from a single draw. Most recommendations suggest no more than 3% of total body volume be drawn per day in patients younger than 2 months of age and no more than 10% in patients older than 2 months (7). This means that approximately 150 mL of blood can be drawn per day from a healthy adult, while only 9 mL per day can be obtained safely from a healthy, 7-pound infant. Smaller amounts are recommended for ill and hospitalized children (8). Table 1 shows the permissible collection volumes as a function of total blood volume estimated from body weight in both healthy individuals and hospitalized patients. In PICUs, blood draws account for 73% of blood loss, and a single patient typically experiences 7.1 ± 5.3 mL of blood loss per day (5,9). This daily loss closely approaches the recommended total loss for an entire month in a hospitalized 7-pound infant.

Collecting the very minimal blood volume required for testing reduces the risk of HAA. The minimum blood volume needed for each test depends on the sum of individual test volume requirements, the volume necessary to measure interference indices, and the amount necessary for an analyzer to accurately pipet a sample (often referred to as the “dead volume”). The dead volume can be significant depending on assay and instrument platform. For example, an assay might require only 30 µL of sample but an instrument’s dead volume might be 10 times greater. Dead volume varies considerably among instruments and also depends on the sample container being used. Table 2 displays the dead volume requirements for various sample cups of two common instrument platforms.

Another volume requirement for chemistry analyzers relates to the measurement of hemolysis, icterus, and lipemic indices. These parameters are usually measured prior to assays on chemistry instruments, and can add approximately 10 µL to the total volume requirement. In addition to these volume requirements, certain instrument platforms have an aliquot storage unit, which stores additional sample in a refrigerated chamber for reflex testing. Such instruments require considerably more volume than recommended minimums and are not a feasible option for laboratories trying to accommodate low sample volume testing.

Small Samples with Big Problems

Most lab result errors occur in the pre-analytical phase of testing owing to specimen collection, specimen handling, and patient variables (10). These pre-analytical errors are behind the high sample rejection rates typically associated with small volume samples. Common small volume sample-related errors include incorrect blood-to-additive ratios, sample evaporation, hemolysis, and human error during manual specimen processing.

Under-filled collection tubes lead to decreased blood-to-additive ratios. Incorrect ratios can cause inaccurate test results, hemolysis, and altered cell morphology. For example, accurate measurement of prothrombin time depends on a specific blood-to-sodium citrate ratio. When citrate tubes are under-filled, the abnormally high concentration of citrate falsely prolongs clotting times. Higher concentrations of additives contribute to other pre-analytical problems, such as hemolysis and changes to cell morphology. Red blood cells exposed to high concentrations of glycolytic inhibitors like sodium fluoride are more likely to hemolyze (11). Likewise, cells   exposed to  hyperosmolar concentrations of ethylenediaminetetraacetic acid (EDTA) may shrink and exhibit an artifactually decreased mean corpuscular volume.

Evaporation also is a major concern for small volume samples given that smaller volumes have a greater surface area-to-total volume ratio. Evaporation in small volume samples can occur during or after sample processing, creating large differences in analyte concentrations. Laboratorians must consider the effects of evaporation before adding on a test. For instance, a sample containing 5 mL of serum (a typical adult volume) that sits open to the air might, after several hours, show an increase in glucose concentration that is up to 10% higher than the initial measurement (7). In contrast, a sample containing only 0.1 mL under the same conditions might show a 50% increase over the initial analysis. To reduce the effects of evaporation, laboratorians should cap samples tightly during storage. Even frozen samples are prone to evaporation when stored for long periods of time.

High surface-to-volume ratios also significantly impact dissolved gases. Labs typically measure total CO2 (TCO2) in plasma via enzymatic techniques following alkali treatment. Alkali treatment quantitatively converts circulating forms of CO2 (HCO3-, H2CO3, and dissolved CO2) to HCO3- , which serves as the limiting substrate for phosphenolpyruvate carboxykinase. At high surface-to-volume ratios, extended exposure of a small volume sample (pCO2 ~40-50 mmHg) to air (pCO2 ~0.3 mmHg) can rapidly deplete the pool of dissolved CO2, reducing apparent TCO2 concentrations. These changes are often incorrectly assumed to reflect metabolic acidosis in the patient, leading to unnecessary additional testing to explain the apparent acidosis. Figure 1 illustrates the magnitude of this phenomenon.

Small volume samples can be collected in reduced vacuum tubes or in microcollection tubes often referred to as “bullets” (Figure 2). These tubes, containing lower quantities of additives, are designed for specimen collection of 0.5-1 mL.  However, using these miniature collection tubes poses several obstacles to laboratory workflow, especially for highly automated laboratories. To begin with, bullets do not fit most analyzers and automated robotic systems. Standard labels also are usually too large for bullets, requiring labs to purchase specialized labels and printers.  Otherwise, labs have no choice but to manually enter patient information into their laboratory information system and analyzers. Adding these manual processes to lab workflows can introduce errors, prolong turn-around times, and increase the number of employees needed.

One potential solution to this problem is to place bullets into larger tubes that can be barcoded and positioned into instruments. Some manufacturers make false-bottom tubes that fit on some instrument platforms, such as the purple top microtainer tube pictured in Figure 2. More often, labs have to transfer small volume samples into a compatible sample cup prior to analysis. Measures taken to accommodate small volumes, such as false-bottom tubes, sample cups, or tube-within-a-tube, all require validation prior to reporting patient results.

Small volume samples are often obtained via capillary blood collection from a finger or heel. Capillary blood collections historically have been used in pediatrics, but this practice is increasingly attractive in adult care. Indeed, the reduced pain (12) and anxiety associated with capillary sampling has been cited as the primary motivation for founding the start-up blood testing company, Theranos. However, dependence on capillary samples may lead to inaccuracy and exaggerated variability. Twenty five years ago research demonstrated that cholesterol measurement in capillary blood is positively biased compared to venous blood (13). A recent study comparing white blood cell (WBC) counts, three-part WBC differential, and platelet counts from successive drops of capillary blood reported the average drop-to-drop coefficient-of-variation was 5 times higher than obtained with well-mixed venous blood (14). 

Capillary samples are also associated with high rates of hemolysis and clotting. Proper collection technique is essential to minimize hemolysis of a sample from a finger or heel stick. “Milking” or squeezing a finger or heel is an absolute non-starter because it can result in hemolysis. In addition, exposing slow flowing capillary blood to disrupted tissue can trigger clot formation. Coagulation testing, therefore, is contraindicated on capillary specimens.

Volumetric Challenges

Drawing minimal volumes often increases the frequency of a lab’s quantity not sufficient (QNS) specimens. In pediatric labs, a common reason for QNS is an elevated hematocrit, which can be as high as 70% in a newborn but normalizes to adult levels by 3 months of age (15). Consequently, newborn samples can yield much less serum or plasma after centrifugation when compared to an identical volume of adult blood which has a hematocrit around 45%. In such cases, more whole blood must be provided so that sufficient plasma is available for analysis. High hematocrit in newborn samples also complicates routine coagulation tests because sodium citrate only distributes into plasma, and not blood cells. In samples with a hematocrit greater than 55%, the resulting plasma citrate concentration is higher than normal, leading to falsely prolonged clotting times. Often this can be corrected by redrawing a sample using a reduced volume of sodium citrate.

Hemolysis also is a very common problem in pediatrics. While not fully understood, this phenomenon may be related to heightened osmotic and mechanical fragility of the neonatal erythrocyte population (16,17). Regardless of mechanism, capillary sampling and the use of small gauge needles undoubtedly exacerbate the problem. Multiple studies have found that drawing blood from a 20-gauge needle or larger helps reduce hemolysis (18,19). However, in children and elderly patients with small or difficult veins, smaller 23- or 25-gauge needles may be required. An analysis of hemolysis rates from various pediatric and neonatal units at St. Louis Children’s Hospital found that the nursery, emergency, and neonatal intensive care units—areas in which heel sticks and small gauge needle use account for the majority of sample collections—had the highest rates of hemolysis. 

Even mild hemolysis may compromise many analyses. Potassium and lactate dehydrogenase released from erythrocytes are notable physiologic interferences that may impact clinical decision-making. In other analyses, the interference is spectral. For example, hemoglobin concentrations as low as 50 mg/dL (Roche Cobas 6000) affect the frequent measurement of direct bilirubin in nursery residents.

Mild-to-moderate hyperbilirubinemia may compromise other analyses as well. Bilirubin concentrations as low as 10 mg/dL routinely observed in neonates affect common enzymatic ammonia assays. Ironically, these ammonia measurements are indicated in patients with liver disease typified by hyperbilirubinemia. Table 3 lists common analytes affected by mild hemolysis and mild hyperbilirubinemia.

Conclusions

Small volume samples introduce several challenges for laboratory testing processes, including numerous pre-analytical, analytical, and post-analytical considerations. To meet these unique requirements, labs must educate staff continually about specimen collection, tube additives, and how minimum volume draws affect repeat or add-on testing. Along with these improved education and communication efforts, new instruments designed to accommodate small samples sizes will be critical in overcoming challenges associated with routine use of small-volume samples.

References

1. Hicks JM. Excessive Blood Drawing for Laboratory Tests. NEJM 1999;340:1690.

2. Koch, C. G. et al. Contemporary bloodletting in cardiac surgical care. Ann Thorac Surg 2015;99:779–84.

3. McPherson, R. A. Blood sample volumes: emerging trends in clinical practice and laboratory medicine. Clin Leadersh Manag Rev 2001;15:3–10.

4. Kurniali, P. C. et al. A retrospective study investigating the incidence and predisposing factors of hospital-acquired anemia. Anemia 2014:634582.

5. Bateman, S. T. et al. Anemia, blood loss, and blood transfusions in north american children in the intensive care unit. Am J Respir Crit Care Med 2008;178:26–33.

6. Linderkamp, O., Versmold, H. T., Riegel, K. P. & Betke, K. Estimation and prediction of blood volume in infants and children. Eur J Pediatr 1977;125:227–234.

7. Jones PM. Pediatric Clinical Biochemistry: Why is it Different? Biochemical and Molecular Basis of Pediatric Disease, 4th Edition: Dennis J. Dietzen, Michael J. Bennett, Edward C. AACC Press.

8. Howie, SRC. Blood sample volumes in child health research: review of safe limits. Bulletin of the World Health Organization 2011;89:46-53.

9. Valentine, S. L. & Bateman, S. T. Identifying factors to minimize phlebotomy-induced blood loss in the pediatric intensive care unit. Pediatr Crit Care Med 2012;13:22–7.

10. Laposata, M. & Dighe, A. ‘Pre-pre’ and ‘post-post’ analytical error: high-incidence patient safety hazards involving the clinical laboratory. Clin Chem Lab Med 2007;45:712–9.

11. Bush, V. The Hemolyzed Specimen: Causes, Effects, and Reduction. BD LabNotes (2003).

12. Woods K, Douketis JD, Schnurr T, Kinnon K, Powers P, Crowther MA. Patient preferences for capillary vs. venous INR determination in an anticoagulation clinic: a randomized controlled trial. Thromb Res 2004;114:161-5.

13. Greenland, P., Bowley, N. L., Meiklejohn, B., Doane, K. L. & Sparks, C. E. Blood cholesterol concentration: fingerstick plasma vs venous serum sampling. Clin Chem 1990;36:628–30.

14. Bond, M. M. & Richards-Kortum, R. R. Drop-to-drop variation in the cellular components of fingerprick blood: implications for point-of-care diagnostic development. Am J Clin Pathol 2015;144:885–894.

15. Dietzen, D. J., Wilhite, T. R., Rasmussen, M. & Sheffield, M. Point-of-care glucose analysis in neonates using modified quinoprotein glucose dehydrogenase. Diabetes Technol Ther 2013;15:923–8.

16. Linderkamp O, Friederichs E, and Meiselman HJ.  Mechanical and geometrical properties of density separated neonatal and adult erythrocytes. Pediatr Res 1993;34:688-693.

17. Goldbloom A, Gottlieb R. Icterus neonatorum. Am J Dis Child 1929;38:57-74.

18. Burns, E. R. & Yoshikawa, N. Hemolysis in serum samples drawn by emergency department personnel versus laboratory phlebotomists. Lab Med 2002;33:378–380.

19. Dugan, L., et al. Factors affecting hemolysis rates in blood samples drawn from newly placed IV sites in the emergency department. J Emerg Nurs 2005;31:338–45.

Khushbu Patel, PhD is a clinical chemistry fellow at Washington University School of Medicine in St. Louis. +Email: kpatel@pathology.wustl.edu

Sarah Brown, PhD, assistant professor of pediatrics and of pathology and immunology at Washington University School of Medicine in St. Louis and of the co-director of the core laboratory at St. Louis Children’s Hospital. +Email: brown_sa@kids.wustl.edu

Dennis Dietzen, PhD is a professor of pediatrics at Washington University School of Medicine in St. Louis and director of the core laboratory at St. Louis Children’s Hospital. +Emaildietzen_d@kids.wustl.edu